“Effort counts twice.”
It is the most memorable idea in Angela Duckworth’s Grit. Talent, multiplied by effort, produces skill. Skill, multiplied by effort, produces achievement. Put the two equations together and effort appears twice. The message is satisfying: whatever advantages you started with, there is something important you can do. [1]
But writing effort into two equations is not evidence that effort has twice the importance of talent. The equations illustrate a way of thinking. They were not estimated from data, and they cannot tell you whether your next hour is better spent practicing, changing strategy, or pursuing a different goal.
That distinction runs through the entire book. Grit is stronger as an argument for sustained, well-directed effort than as a scientific explanation of exceptional achievement. It offers useful advice about practice, interests, goals, and environments. It is less convincing when those ideas become a broad theory built around a single personality score.
My verdict: read the book for ways to improve your work. Be more skeptical when it tells you what kind of person succeeds.
The important question is not whether persistence matters. Of course a long project generally requires someone to keep working on it. The harder questions are how much a grit measure adds to what we already know, whether increasing it causes better outcomes, and when persistence stops being the right response.
Those are the questions that should determine how seriously we take the theory.
A fair reading is better than an easy takedown
Duckworth’s argument is more thoughtful than “try harder and anything is possible.” She distinguishes talent from developed skill, and skill from productive accomplishment. She describes interests as something people can cultivate. She emphasizes feedback, supportive adults, culture, and the possibility of changing lower-level goals. [2]
She also makes concessions that critics should acknowledge. Opportunity can matter more than individual psychology. Talent differences are real. Rest belongs in demanding practice. Hardship can damage people. Grit is not the highest moral virtue, and staying with every failing project would be foolish.
The endnotes add further qualifications. They identify unpublished evidence, explain adaptations to the questionnaire, and acknowledge disputes about deliberate practice and the limits of some interventions. This is not a book with no scientific documentation. In several places, the documentation is more careful than the impression left by the main narrative. [3]
The strongest version of its thesis is worth taking seriously: people can improve the practices and conditions that allow them to work effectively over time.
The stronger version needs more support: grit is a distinctive, unusually powerful general explanation of achievement, and the book’s recommendations reliably increase it in ways that improve life across domains.
Those are different propositions. Much of the controversy comes from moving between them as though the evidence for the first automatically proves the second.
The three claims behind the score
This review separates the book’s central argument into three parts.
Grit predicts achievement and sometimes matters more than talent. The research contains genuine predictive signal, but the answer changes with the population, outcome, and comparison. The broad superiority claim receives a 50% rating.
Sustained, high-quality effort develops skill and helps turn it into accomplishment. This is the strongest proposition, rated 75%. Useful practice is real; the two equations are a conceptual illustration rather than a quantitative law.
The recommended assets and environments can build grit and improve long-term success. Some interventions produce meaningful results. Evidence for a durable, general-purpose grit-building recipe is much weaker. This claim receives 50%. [4]
Together, those judgments produce a Scientific Accuracy score of 58%. The point of separating them is to avoid a useless argument in which either every persistence finding validates the whole book or every limitation makes persistence irrelevant.
The evidence supports a more interesting conclusion: important parts of the practical advice can survive even when the grand explanation needs to shrink.
Staying in the program is one kind of success
The opening West Point research asks an interesting question: which apparently qualified cadets will remain through a demanding early training period?
Measuring a characteristic before the outcome is a real strength. It is more informative than interviewing successful people afterward and finding that they describe themselves as persistent. The grit research provides evidence that a persistence-related measure can predict retention in particular demanding settings. [5]
But staying and excelling are different outcomes. A measure that helps predict whether a person completes initial training may be less useful for predicting academic performance or later excellence. The book recognizes that admission measures can predict subsequent performance even when they do not distinguish early leavers. Later West Point research also supports outcome-specific contributions. [6]
There is another complication: the cadets were selected before the study began. An admissions process restricts the range of some characteristics. Comparing people who already cleared a demanding threshold cannot establish how little the threshold-related abilities matter in the wider population.
Imagine studying only excellent pianists and finding that rehearsal habits predict who performs most consistently. That would be useful. It would not demonstrate that musical ability barely matters among everyone who might sit down at a piano. The comparison group does part of the work.
Retention also has an ambiguous relationship with personal welfare. Staying can be good. Leaving can reflect poor fit, an injury, a better opportunity, or a sensible revision of plans. An institution may want lower attrition while an individual benefits from departing.
The practical lesson is to name the outcome before celebrating a predictor. “This helps forecast early completion here” is a useful claim. “This explains who succeeds in life” is a much larger one.
Grit has a novelty problem
Grit combines perseverance of effort with consistency of interests. The obvious comparison is with established personality traits, especially conscientiousness.
In a major synthesis of 88 samples involving 66,807 people, the observed correlation between grit and conscientiousness was .66. A correction for measurement error produced an estimate of .84. Those are large overlaps, although the corrected figure depends on its assumptions and is not the raw correlation between individual scores. [7]
Overlap alone doesn’t make a measure useless. A narrower scale can be convenient. It can highlight a useful facet. But a new name, a new questionnaire, a new theory, and a useful new prediction are four different accomplishments.
The relevant test is whether grit adds meaningful information beyond good measures we already have. In the adolescent twin research discussed in the appendix, the Big Five explained 5.6% of variation in GCSE performance. Adding the grit facets increased the explained variance by about half a percentage point in that setting. [8]
That is additional information, but it is a long way from discovering the dominant missing ingredient in achievement.
This is why I would be cautious about using a grit score in hiring or admissions. A statistically detectable relationship is only a starting point. You would still want evidence that the measure improves decisions in your setting, predicts the right outcome, and adds enough value to justify the mistakes it will make.
Duckworth herself warns about the ease of faking high-stakes self-reports. The book’s ten-item questionnaire is documented as an adaptation of the original form, but that documentation doesn’t turn it into a current, purpose-specific selection instrument. [9]
Comparison standards matter too. In later large studies, grit self-reports could predict outcomes positively within a school while giving a misleading comparison between schools. A demanding setting can make a capable student judge the same behavior more harshly. That is a reason to check the setting and comparison—not to conclude that every grit association is worthless. [26]
Use the questionnaire to start a conversation about behavior. Don’t let the neatness of a score persuade you that you have measured someone’s future.
“Passion” is doing two different jobs
In ordinary language, passion suggests enjoyment, fascination, or intense commitment. In the grit measure, it is largely represented by the stability of interests over time.
Duckworth explains this usage. The issue is not that the definition is hidden. The issue is whether interest consistency adequately captures the richer qualities invoked by the surrounding story. You can keep pursuing something you dislike. You can change direction because a new activity fits your abilities and values much better. [10]
A later synthesis of academic studies reported average correlations of about .19 for overall grit, .21 for perseverance, and .08 for consistency of interests. The sample was large: 137 studies, 156 dependent samples, and 285,331 participants. The figures suggest that perseverance carried more of the average academic signal than stable interests did. [11]
These are associations, and they should stay associations. They don’t tell us that teaching someone to persist will create exactly the corresponding improvement in grades. They do make it harder to justify treating the two facets as equally informative expressions of one powerful general ingredient.
Later work tried measuring attained passion separately, but its interpretation is contested. In a 2019 exchange, critics questioned treating the whole short grit scale as perseverance and did not recover the perseverance-by-passion interaction when they analyzed its facets separately. One interest-consistency interaction remained for grades. The original researchers replied that the apparent facets were wording artifacts and defended their approach. Their reply also reported tests of reworded scales with new participants. This is a measurement and reanalysis dispute, not a retraction or a new replication of the original performance interaction. It gives us another reason to keep the questions separate. [27]
For a practical decision, I would separate the questions. Are you putting in effective effort? Do you enjoy the activity? Does the goal still matter? Are you learning? Are you continuing because the project is promising, or because changing would feel like a confession?
Those questions can point in different directions. A single grit score can flatten them.
The book’s discussion of developing interests is therefore more useful than a literal rule to keep interests unchanged. Explore, gain experience, and examine the actual work. Consistency can follow a good match. It need not be imposed before you have found one.
The effort equations explain an idea, not a result
In chapter 3, “Effort Counts Twice,” Duckworth presents two equations:
Talent × effort = skill.
Skill × effort = achievement.
The two-stage idea is reasonable. Work can increase your capabilities, and using those capabilities can produce something valuable. A skilled potter who makes no pots contributes differently from a skilled potter who keeps working. [1]
The trouble starts when the algebra is treated as evidence about relative importance. Substituting the first equation into the second produces a squared effort term because the same symbol was used twice. It tells us nothing about whether those two kinds of effort have the same meaning, scale, or returns.
Training effort and performance effort can be different activities. Practice quality matters. Prior knowledge matters. Fatigue can change what another hour accomplishes. The equation gives you no estimate for any of these quantities.
This matters because the slogan offers an answer before the practical problem has been diagnosed. Perhaps you need more repetitions. Perhaps you need better instruction. Perhaps the market does not value what you are making. Doubling the wrong activity is an expensive way to discover that effort was not the limiting factor.
I would keep the conceptual distinction and drop the implication of a measured law. Effort contributes at multiple stages. Its marginal value depends on the task and the circumstances.
The useful question is therefore concrete: what would make the next unit of work more productive? That question can lead to more effort, but it can also lead to better feedback, a changed approach, or a different problem worth solving.
Practice is the strongest practical chapter
The book is excellent at distinguishing time served from deliberate improvement.
A person can spend years repeating the same comfortable activity without systematically addressing weaknesses. Duckworth instead emphasizes a specific improvement target, concentrated work, useful feedback, and revision. She also includes rest and treats the famous 10,000-hour figure as a rough average rather than a universal threshold. [12]
That is valuable advice. It would remain valuable even if the word “grit” disappeared from the cover.
The uncertainty concerns the strongest explanatory claims. Research linking practice histories to performance has difficulty separating practice from prior ability, opportunity, instruction, motivation, and the way practice was measured. The cross-domain findings also vary. A synthesis reporting different practice–performance relationships in music, sports, education, and professions cannot be reduced to one rule that explains exceptional achievement everywhere. [13]
Nor should the unexplained variation be relabeled “talent.” That would repeat the same error in the other direction. A statistical remainder contains many possibilities.
The practical distinction is between a learning loop and a total theory of success. In a well-defined skill, you may be able to identify an error, practice a correction, and check whether performance improves. In entrepreneurship or research, the problem and feedback can be much less clear. Working harder inside a mistaken model can reinforce the mistake.
For a reader, I would translate the chapter into one question: what are you getting better at, and how will you know?
An answer like “I spent three hours on it” is incomplete. An answer that identifies a capability, a specific weakness, and a credible measure of improvement gives effort somewhere useful to go. That is the chapter’s strongest contribution.
A program can help without validating a personality theory
There is real intervention evidence in this literature. Dismissing all of it would be as careless as treating every positive program as a triumph for grit.
One set of classroom experiments reported a roughly 0.2-standard-deviation improvement in a standardized math test at a 2.5-year follow-up. That is a meaningful favorable result for a particular educational program. It should receive credit. [14]
But the intervention was a program, not a one-unit increase in an abstract personality trait. A curriculum can change study behavior, expectations, teacher interactions, or other features of learning. Its label does not identify the active ingredient.
A later Swedish program combined deliberate practice with growth-mindset instruction. It increased deliberate-practice behavior on a test but did not improve the measured attitudes or mathematics performance. Teachers randomized mentor groups rather than individual students, and the program’s delivery differed from the earlier interventions. The mixed result limits a broad transfer claim; it does not show that targeted practice cannot help. [28]
Other programs produce different patterns. A 2026 working paper from a trial across 100 English schools reported a sizable shift in beliefs about whether intelligence could change, but no detected effect on national achievement tests. This later evidence informs how we should read the book now; it is not something a 2016 book could have cited. [15]
A belief can move without a performance measure moving with it. That is exactly why measuring the intended outcome matters.
The purpose example supplies a smaller version of the same lesson. In the study discussed in the appendix, students spent about 49 rather than 25 seconds per review question. The near-doubling is real for that task. It is not evidence that students generally studied twice as much or became twice as gritty. [16]
The book also describes job crafting at Google: redesigning parts of your own work to make it more meaningful. The final report of the earlier corporate work describes workshop-level condition assignment, not individually randomized employees. The job-focused group showed a tentative rise in happiness at six weeks. Statistical support for that result was less clear after adjustment for multiple comparisons, and the study did not establish a lasting gain at six months. Combining changes in views of the self and the job showed later happiness benefits, but not sustained performance gains. A separate randomized study supported some later happiness benefits, with substantial follow-up loss. These results leave room for useful changes at work without establishing that purpose creates durable grit. [29]
The right conclusion is conditional: some practices and programs can improve some behaviors and outcomes. That is enough to justify further use and testing where the evidence fits. It is not enough to claim a reliable process for raising a single durable trait that improves school, work, parenting, and life together.
The environment can change while the score stays put
The culture chapter describes reduced attrition at West Point even though incoming self-reported grit was approximately unchanged. It is a historical account rather than a controlled experiment, but it usefully redirects attention toward the setting. [17]
What changed in training, support, expectations, or the experience of being there? Those questions may be more useful than asking why a new generation failed to arrive with better character.
The same principle applies to parenting and extracurricular activities. A child may persist because an activity is well taught, accessible, enjoyable, and socially rewarding. Participation can reflect these conditions as well as the child’s prior tendencies. Observing later success among participants cannot, by itself, tell us how much a general persistence trait was created by the activity. [18]
Duckworth acknowledges this selection problem. She also recognizes unequal access to good opportunities. Those are important strengths of the book.
Where I would narrow the advice is in the promise of broad transfer. A sport can be worthwhile because the child learns the sport, enjoys teammates, and develops specific skills. Those direct benefits do not need a promise that the activity will manufacture grit for every other part of life.
Her family’s Hard Thing Rule is similarly best understood as a proposal to consider, not a tested dose of character development. It includes choice and a natural stopping point, making it more reasonable than “never quit.” But its particular package was not validated in the material reviewed here. [19]
For a parent or manager, the productive move is to improve the conditions for effective effort. Clear expectations, usable feedback, real support, and room to report problems give persistence a better chance to accomplish something worthwhile.
Even a credible feedback message depends on context. A later undergraduate experiment found no trust advantage from wise feedback when both groups already had high trust. That is a boundary on transfer, not conclusive disproof of the original school experiments about racial mistrust—and neither study validates an entire parenting-for-grit formula. [30]
Persistence needs an exit rule
A theory of perseverance has to handle the person who should stop.
Duckworth’s distinction between high-level goals and lower-level tactics helps. You can stop using a bad tactic while remaining committed to a meaningful purpose. Changing a product is not necessarily abandoning the ambition to build a useful business. [20]
But the flexibility can also make the theory hard to test. A person succeeds by persisting, so grit gets credit. A person succeeds by switching, so the switch is described as persistence toward a higher-level goal. Unless the levels and stopping conditions are specified beforehand, the theory can absorb opposite behaviors after the fact.
A practical rule needs to risk being wrong. What result would justify continuing? What evidence would change the plan? What cost would make even the broader goal no longer worth pursuing?
The research on costly perseverance provides a reason to ask these questions. Some laboratory findings associate higher grit with persistence that carries costs. These results do not show that gritty people generally make bad decisions. They show why more persistence should not automatically be treated as better. [21]
Here is the distinction I would use: be loyal to the reasons for a goal, not merely to the fact that you once chose it.
Health, family, ethics, money, and alternative opportunities belong in the decision. They are not distractions from the true score called achievement. If reaching a target destroys what made the target worthwhile, successful persistence has become a strange kind of failure.
The book recognizes several of these concerns. Its practical guidance would be stronger if they were built into a clear decision rule rather than left for the reader to assemble.
Hope is useful when the comparison stays clear
The chapter on hope brings together several different ideas: interpretations of setbacks, growth mindset, learned helplessness, and psychological treatment. The connection is understandable. Each offers a way to think about how people respond when effort becomes difficult. But the evidence for one cannot automatically certify the others. [22]
The book’s endnotes already acknowledge that the Penn Resiliency Program did not outperform active comparison conditions in a synthesis. A later Dutch trial of an adapted, universal school program also found no depressive-symptom benefit at one or two years. Different populations and delivery matter; neither result erases every favorable prevention finding. [31]
The cognitive-therapy example shows why the comparator matters. The follow-up evidence discussed in the appendix supports enduring benefits relative to withdrawal from medication among treatment responders. That is a different comparison from showing that therapy universally has longer-lasting benefits than continued medication. The treatment stage changes the question. A later synthesis supports enduring benefits after treatment ends, but that still does not establish universal superiority over continued medication. [23]
Duckworth also encourages seeking professional help rather than expecting a short chapter to undo entrenched problems. That is an important safeguard, and this review is not advice to change treatment.
The wider practical lesson is to keep hope connected to available action. A more constructive interpretation of a setback can help someone try a new strategy or seek support. It does not establish that the original goal is achievable under every constraint, or that optimism is the missing ingredient in every failure. Sometimes a hopeful next step is changing the situation rather than changing what you believe about it.
What I would take from the book
Start with a worthwhile goal and inspect the actual work it requires. Are you interested in doing that work, or only in being the person who has the result? Those are different motivations, and the distinction can save a great deal of wasted effort.
Build a feedback loop around a specific capability. Practice something that matters, get useful information, and revise. Look for support that makes the work sustainable rather than treating every need for help as a weakness.
Then decide in advance when to review the commitment. The review should include progress, opportunity costs, and whether the original reasons still hold. A plan without a review point can quietly become an identity test: continuing proves resolve, while leaving feels like a verdict on character.
Imagine a founder whose customers repeatedly decline a product. Persistence might mean improving the pitch, changing the product, choosing a different customer, or ending the project. “Be grittier” cannot select among those actions. Evidence about the problem can.
This is the spirit in which I would use the book’s goal hierarchy: stable enough to support sustained work, flexible enough to learn. The distinction between a goal and a tactic is valuable. Even an important goal should remain open to revision when its justification changes.
Keep: deliberate improvement, feedback, developing interests, supportive environments, and the separation of purpose from tactics.
Modify: broad causal claims, the treatment of interest consistency as passion, and prescriptions that imply the same pattern of persistence is best across domains.
Reject: using a popular grit score as a verdict on someone’s potential or using the effort equations as a measured ranking of causes. [24]
These are practical judgments drawn from the review, not a separately tested program. The objective is to make effort more productive and better directed, not to replace one branded route to success with another.
Why the rating is 64%
Scientific Accuracy: 58%. The book has real research behind it, but a large gap remains between context-specific prediction, useful practices, and the strongest claims about a distinctive trainable trait.
Reference Accuracy: 83%. The complete endnotes are a substantial strength. They document important qualifications and competing evidence. Good sourcing can coexist with a main argument that travels further than the sources warrant.
Practical Value: 50%. Selected recommendations are useful. The broader grit-building promise, transfer across contexts, and operational guidance about when to disengage remain insufficiently established. [25]
The overall percentage averages these three judgments. It is not a measurement of how much of your success is under your control, and it is not a score for Duckworth’s character or competence.
Grit is worth reading selectively. Its best advice is to work thoughtfully on things that matter. Its weakest implication is that a single personal quality explains who gets there.
Effort matters. Choosing where it goes matters too.